Data Ethics — Beyond Compliance

Data Ethics — Beyond Compliance

Ethics ≠ legal. DMBOK's ethics chapter: principles, risks, and the ‘should we even build this?’ gate.

Why it matters

Legal compliance is the floor; ethics is the ceiling. A dataset can be lawful to collect, retain and infer from — and still cause harm. DMBOK's ethics chapter (§2 in DMBOK²) names the principles before the lawsuit names them for you.

Going deeper

DMBOK's ethical-data principles (paraphrased):

1. Impact on people first — does the use harm the data subjects?
2. Potential for bias — does the data over- or under-represent groups?
3. Transparency & accountability — can affected people see how their data is used?
4. Risk reduction — proportional safeguards relative to harm potential.
5. Cultural impact — second-order effects on society / workforce.

An ethics review gate — a 30-minute conversation before sensitive use cases — is the cheapest investment with the highest avoided-harm payoff. It's the data equivalent of an IRB.

Data ethics is the line between ‘legal’ and ‘would I be comfortable on the front page tomorrow?’.

A pharmacy that legally sells every product on its shelves to anyone with a prescription is compliant. A pharmacist who refuses to fill a prescription that would clearly harm the patient — even if technically legal — is ethical.

Data teams face the pharmacist's choice constantly: ‘we can collect that, we can infer that, we can sell that’. The ethics review is asking ‘should we?’ before the lawsuit, the regulator, or the journalist asks.